This is probably an unpopular opinion on HN but I don't think patents should be transferable. If only the inventor were allowed to benefit from a patent it would work closer to the original intent of the patent system which was to incentivise creators. Since then it's purpose has been perverted into a right to own IP. Ownership of IP only benefits lawyers and patent trolls. Please note that I'm not talking about trade secrets which are another thing altogether.
The inventor can benefit from a patent by selling it.
Maybe they don't want to raise millions of dollars and spend years building a business that may or may not eventually generate $20 million in profit over the next couple decades, and would rather just sell it now for $1 million and get back to creating!
Even if they wanted to take the time and risk to use it directly, the skills required to invent something and the skills to profit from it aren't necessarily the same.
> Maybe they don't want to raise millions of dollars and spend years building a business
But see, this is the valuable part! This is what we should be rewarding! The idea is worthless without execution. The concept that someone who simply had a relatively trivial idea should be able to extract rent from someone else who had the same idea but actually spent years executing on it and building a real business that benefits actual customers, taking on significant risk to do so, is repugnant.
My uncle have invented a new type of sprinkler. Literally think of bearded guy tinkering in his garage for years. I can't imagine why his idea would be worthless without execution. Instead, this is something I very firmly believe should be the focus of the patent system. He made some money on it and couldn't have without the patent system. Executing this idea is quite capital intensive as you'd need to build a factory and so forth. It's much more effective to just sell the rights to an existing manufacturer.
Of course this is an idea resulting in an actual physical piece. If you mean that patents not resulting in such shouldn't be allowed ... then yes, I might agree.
Patents are supposed to do two things (as far as I'm concerned): Encourage invention and reward the inventors.
I think your uncle is the kind of guy who would be working on this stuff regardless of what reward is on the other side, and my experience is that most inventors are like that. Open source software has shown that all inventors really need to work hours and hours on developing new things, is their own drive and maybe some recognition. So I don't think patents are at all necessary for encouragement.
But we all agree he deserves reward. What you're assuming is that patents is the only way that could happen. I don't believe that's the case. I believe patents is locking us into thinking about patents as the only method of reward. If we abolish it then other, more productive methods will become prevalent. I'm talking about research grants, competitions, award prizes, investments, or employment. People will demand it when patents are absent, and politics will shift to encourage it.
I think the only way patents can be redeemed, is if they're non-transferrable and the inventor is forced to license it for a fair price within a year of registering. There also needs to be more legal aid to help small scale inventors enforce their patents. There are so many things that need to be fixed about patents.. but the thing is they're just fundamentally flawed because they're based on the idea of a monopoly on ideas. Monopolies are always a bad idea unless they're controlled by a democratic entity.
> I think the only way patents can be redeemed, is if they're non-transferrable and the inventor is forced to license it for a fair price within a year of registering.
If the inventor can't transfer the patent but can still license it then there will immediately be lawyers offering to "find" patent licensees in exchange for a share of the license fee and we'll be right back where we started. You're also doing nothing to solve the nearly intractable problem of what a "fair price" for a patent is.
The fundamental issue is that in practice patents aren't actually used by small inventors on any sort of meaningful scale. They're too expensive to get and way too expensive to enforce for that. But you can always find that one anecdote to trot out in favor of a policy that in reality primarily benefits the likes of IBM.
But let's show the counterpoint with some counter-anecdotes. Think of any famous rags to riches (or middle class to riches) tech star. Larry Page, Mark Zuckerberg, anybody. Even the ostensibly pro-patent ones like Bill Gates or Larry Ellison. Now how many of them made most of their fortunes inventing things and then licensing the patents and how many did it by inventing things and then making and selling them to the public? Exactly.
Because patents are used to steamroll over the little guy, not help him. They're a thing that can only be wielded by lawyers and large companies with lawyers, which means they always put the little guy at a distinct disadvantage regardless of whether the little guy is the inventor or the implementor or both.
So instead you have a company that gathers up contracts with owners of IP from a bunch of places to get a "finder's fee" from anybody they negotiate a patent license with. It's no different at all.
About the only thing you could even hope for is that the original inventor will have more of a conscience about how the patent is being licensed than the troll entity. But if that happens on a regular basis then then the troll entity will just start putting liquidated damages clauses in their license finding contracts with inventors that penalizes the inventor for pulling out after the troll entity has done work to find a license victim.
As someone who's spent a ton of time in research labs and has gone halfway through the patent process a couple times - no, no it isn't. A lot of people think this and fail to recognize how hard it is to come up with something legitimately new and verify that it works. Ideation and prototyping are fundamentally different skills than formalization and "execution" and demanding t the same person does both is wasting multiple separate talents.
> fail to recognize how hard it is to come up with something legitimately new
The vast majority of process and software patents seem to not actually be new though, and the patent system explicitly doesn't reward 'new' in the general sense, it rewards 'new to the patent system' (aka first to file)
> and verify that it works.
In most patent troll litigation, the holder of the patent hasn't verified or actually built anything.
In my opinion, research labs should be rewarded for their work finding genuinely 'new' things. That includes benefiting from another entity building or developing what they found. It doesn't include sitting on the patent waiting for someone else to 'discover' or simply do the same thing, wait for them to profit for a few years, then sue.
I agree with what you've said here: inventing things that are actually new is difficult and the people that don't invent things that are actually new shouldn't be rewarded. Patent trolls are a thing, and people patenting boring junk is also a thing. The things is, saying things like "only execution matters" goes too far in the other direction, and, like you said, cuts out research labs and similar endeavors.
> As someone who's spent a ton of time in research labs and has gone halfway through the patent process a couple times - no, no it isn't.
"Worthless" is an exaggeration. The fair point is that implementing the idea (and all of the supporting infrastructure necessary to do it) is generally much more work than coming up with it to begin with.
So you spend 250 hours to come up with an invention and somebody else spends the same 250 hours to come up with the same invention but also spends 5000 more hours to implement it and then the patent system says you can go to them and say "that's mine" and claim all the profits.
On the other hand, if you don't have the resources or skills to implement, but do have the (different, and often rare!) skills necessary to ideate, it'd be just as clear an injustice if you couldn't get anything for your inventions, or if you were incredibly good at inventing but were forced to waste 99.9% of your time executing using engineering skills that might as well have been outsourced to India. The system needs to support garage-workshop inventors and academic theoreticians just as much as it needs to support Intel.
> So you spend 250 hours to come up with an invention and somebody else spends the same 250 hours to come up with the same invention but also spends 5000 more hours to implement it and then the patent system says you can go to them and say "that's mine" and claim all the profits.
The problem here is the "all the profits" part, because that's just as extreme in the opposite direction. How about a solution where the 250-hour inventor gets a portion of the profits proportional to the time and effort put into it by each party, say a perfectly linear 250/(250+5000)? Then the inventor doesn't get much profit out of it, but because he spent less time on it he can theoretically do this repeatedly and make a reasonable amount of money. Which is, really, how the modern patent system would work if patent grants were done optimal, which is admittedly impossible.
> On the other hand, if you don't have the resources or skills to implement, but do have the (different, and often rare!) skills necessary to ideate, it'd be just as clear an injustice if you couldn't get anything for your inventions, or if you were incredibly good at inventing but were forced to waste 99.9% of your time executing using engineering skills that might as well have been outsourced to India.
It wouldn't. The primary cause of the injustice is that someone can spend thousands of hours implementing something before they find out someone else can claim the rights to it. The would-be inventor would know there is no patent waiting for them from the start.
If the only ideas you're capable of having require more resources to implement than you have then you would go work for someone with those resources and collect a salary. Which is what happens in the overwhelming majority of cases regardless of patents.
> The system needs to support garage-workshop inventors and academic theoreticians just as much as it needs to support Intel.
Which it doesn't, and can't. The garage-workshop inventor doesn't have the resources to litigate a patent. At best they can sell it to someone who does for pennies on the dollar and thereby get ripped off, which is why hardly anybody does that. Most inventors in practice either implement and sell their invention themselves or are the employees of someone who does. Or do grant-funded research which the public paid for and should therefore get.
> The problem here is the "all the profits" part, because that's just as extreme in the opposite direction. How about a solution where the 250-hour inventor gets a portion of the profits proportional to the time and effort put into it by each party, say a perfectly linear 250/(250+5000)?
Then you end up rewarding incompetence and inefficiency. The slower and more wasteful an inventor is the more they get paid. And you're creating a huge incentive to lie about things that are hard to prove. How does anybody else know how long it took you to do something?
This isn't absolutely true, if ideas are worthless without execution then corporations would never spend millions and millions inside their R&D departments creating inventions which may or may not come to market.
The assumption that ideas are always trivial isn't correct, sometimes the inventor spends years and years perfecting a design in his own garage, but doesn't have the means to bring it to market. Should the inventor not get rewarded for his work? Nothing is wrong with the final invention nor is it from the lack of trying to bring it to market, he just doesn't have the financing necessary to pull it off.
Should the inventor not get rewarded for his work?
The patent isn't a reward for his work, though. The patent is a reward for having an idea and filing for it. The inventor could just as easily spend years in the garage and come up with nothing patent-able, thereby obtaining no reward.
Sans patents, BigCos are forced to compete in a commoditized market even more cutthroat than generic drugs (since those are at least regulated by the FDA).
Society isn't rewarding it, as it's not the government who buys the patent from the inventor. The patent system just lubricates a market that would still happen as long as NDAs are legal.
If we ignored the value of the idea, then those with the means to scale up production quickly could easily steal ideas from those who are trying to find the means to scale up production.
There are several things that go into a product. Using your metric, if you chip away any one of them the other parts tend to become worthless.
You wouldn't have a product without the ideas that go into it. You wouldn't have a product worth anything if you didn't have people selling & marketing it.
In a vacuum each aspect is worthless. So programmers should be paid zero, just like inventors should be paid zero. Because unless you're going to build a business, what you just programmed is worthless.
I'm sure a network of experts in IP management would step up in such a hypothetical world, and these people would be more than happy to 'build the business' around patents on behalf of the inventors, in exchange for money of course. Pretty similar to the current system on the face of it. But the nuances may have some greater merit.
This. A good patent is a valuable thing and I can't get behind barring people from selling valuable things that they made or that they bought themselves.
Whenever PatentTalk comes up we see a lot of Band-Aid solutions like this. Similarly with copyright issues. Instead of trying to hack around the shitty laws and regulations we have in place for this stuff, let's just fix the problem at its source. In the case of patents, this mostly means to quit handing out bogus patents, and making it much easier for a patent to be invalidated in the first place. Especially with the sort of patents floating around nowadays, not to mention the litigiousness, a patent like this being invalidated should not be 'big news' - it should be a routine occurrence.
If the inventor is a corporation, this would be gamed by simply transferring ownership of the corporation. If the inventor is a human, the person would have to organize as a sole proprietorship and forgo limited liability. Designing rules is hard!
It's fine for corporations to own patents, they should be however nontransferable so only that corporation can ever bring a lawsuit to enforce it. And if the corporations is bought or sold or dissolved or rearranged, it loses the patent. That sort of limits the scope of how the patent can be used. Which would still be an improvement than now.
That would mean that any IP-heavy company could not get any investment, as selling any significant stake would mean all of their patents would be invalidated.
Making patent ownership non-transferable would have no significant effect. Patent troll companies would simply contract with the inventor and sue "on behalf of" the inventor. Same dollar amounts would change hands, the contracts would just be written differently.
One troll is already famous for doing exactly that. Just can't recall his name. Company brings in huge money. I do remember he always warns they can take the offer or he will "go thug" with lawyers.
One practical reality of this rule would be that only wealthy inventors could secure effective IP rights. A patent without the ability to finance a lawsuit (about $2 million, min) is an empty threat, and sophisticated parties know this and will refuse to license your IP. Contingency fee lawyers used to fill this void, but recently patent suits have become so unprofitable that, at least anecdotally, very few firms are taking contingency cases. Private equity will step in occasionally for patents with a pedigree, in a strong portfolio, but it's rare.
There is an analog with art, a /droight de suite/ or a system for resale royalties. The idea is that starving artists have very low bargaining power compared to galleries and art dealers. The theory goes, that if the artists are disallowed from selling all of their ownership rights, that is, they will receive a royalty every time their painting is sold in the future, then the creators of pieces that turn out to be extremely valuable will never be left starving on the street simply because they were bad at negotiating.
However, retrospective analysis of such systems in practice has tended to show that artists' bargaining position has not been as weak as the common narrative, and constraining the upside potential for art dealers has lead to less funding in the market overall, and less risk-taking in the form of purchasing works from undiscovered artists who are most likely to resemble the starving artist archetype.
> closer to the original intent of the patent system which was to incentivise creators
Speaking about incentives, the USPTO still has an incentive to accept as many patents as possible. This is a bit like paying fishermen for the fish they didn't catch.
My opinion is probably even more unpopular. I don't believe software should be patentable at all.
Algorithm implementation is already covered by copyright and that is enough.
The American patent is working backwards. Instead of fostering innovation, it makes inventive people stop to wonder if what they are doing is or not infringing any of so many bogus patents.
I think patents should be transferable, I just think patents should be a lot harder to acquire to the point where they are rare, which was absolutely the original intent (at least as it pertains to US patents).
See, for example, Jefferon's letter to Isaac McPherson:
If Jefferson were reanimated now he would, pretty clearly, be livid about the current state of the patent system and the way patents are essentially rubber-stamp granted as long as you pay the fee and use all the right legalese to make your incredibly obvious "this, but on a computer" invention seem like it is anything new.
Absolutely and the same can be said for the original idea behind copyright. Heck that descendants can live off royalties of a passe one's copyright is not why it was invented.
I would like to see one transfer. It would have to be worded just right, but just one transfer.
The reason I call for one transfer is patenting someting is a big deal(financially) for a poor person. The fees are too high. (I actually contacted the patent office in the US a few years ago, and questioned their fees. They came back with we give a slight break to micro-businesses, or someting like that? It was not low enough for my budget.)
My point is a rich entity can patent a fart sound, and the cost is a two person corporate lunch. And that's why they patent everything?
I would like to see patents/trademarks costs go up in cost exponentially. Example, first patent $1000 dollar fee. Tenth patent $50,000? I would like to see the cost tied to income level, maybe even gross?
A poor entity, even if they fill out the paperwork themselfs; the cost of a patent is high. Especially when you have file multiple patents in order to protect your idea. That person should be able to profit on his/her idea? That person should be able to afford to protect their idea. So many poor people can't afford those patent fees--
In all reality, I would like to see patent/trademark fees lowered for poor entities. (If the entity sends in sloppy paperwork; you can make that poor person pay what a rich person pays. This would prevent abuse of the system?)
The company you work for on whose time you invent something that is granted a patent should give a revocable license to your employer but the inventor should never have to give up the patent to the employer. Unfortunately that's how it's handled in most places and a reason why many do not submit stuff to patent lawyers in the company. I'm curious how this works in the US with IBM creating at least one patent a day.
